mteb 2.6.4__py3-none-any.whl → 2.6.6__py3-none-any.whl

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (142) hide show
  1. mteb/abstasks/classification.py +2 -3
  2. mteb/abstasks/multilabel_classification.py +3 -3
  3. mteb/abstasks/regression.py +1 -1
  4. mteb/abstasks/retrieval.py +1 -1
  5. mteb/abstasks/task_metadata.py +9 -14
  6. mteb/descriptive_stats/Retrieval/NanoClimateFEVER-VN.json +30 -0
  7. mteb/descriptive_stats/Retrieval/NanoDBPedia-VN.json +30 -0
  8. mteb/descriptive_stats/Retrieval/NanoFEVER-VN.json +30 -0
  9. mteb/descriptive_stats/Retrieval/NanoHotpotQA-VN.json +30 -0
  10. mteb/descriptive_stats/Retrieval/NanoMSMARCO-VN.json +30 -0
  11. mteb/descriptive_stats/Retrieval/NanoNQ-VN.json +30 -0
  12. mteb/descriptive_stats/Retrieval/TVPLRetrieval.json +30 -0
  13. mteb/models/model_implementations/align_models.py +1 -1
  14. mteb/models/model_implementations/andersborges.py +2 -2
  15. mteb/models/model_implementations/ara_models.py +1 -1
  16. mteb/models/model_implementations/arctic_models.py +8 -8
  17. mteb/models/model_implementations/b1ade_models.py +1 -1
  18. mteb/models/model_implementations/bge_models.py +45 -21
  19. mteb/models/model_implementations/bica_model.py +3 -3
  20. mteb/models/model_implementations/blip2_models.py +2 -2
  21. mteb/models/model_implementations/blip_models.py +8 -8
  22. mteb/models/model_implementations/bmretriever_models.py +4 -4
  23. mteb/models/model_implementations/cadet_models.py +1 -1
  24. mteb/models/model_implementations/cde_models.py +2 -2
  25. mteb/models/model_implementations/clip_models.py +3 -3
  26. mteb/models/model_implementations/clips_models.py +3 -3
  27. mteb/models/model_implementations/codefuse_models.py +5 -5
  28. mteb/models/model_implementations/codesage_models.py +3 -3
  29. mteb/models/model_implementations/cohere_models.py +4 -4
  30. mteb/models/model_implementations/colpali_models.py +3 -3
  31. mteb/models/model_implementations/colqwen_models.py +8 -8
  32. mteb/models/model_implementations/colsmol_models.py +2 -2
  33. mteb/models/model_implementations/conan_models.py +1 -1
  34. mteb/models/model_implementations/dino_models.py +19 -19
  35. mteb/models/model_implementations/e5_instruct.py +23 -4
  36. mteb/models/model_implementations/e5_models.py +9 -9
  37. mteb/models/model_implementations/e5_v.py +1 -1
  38. mteb/models/model_implementations/eagerworks_models.py +1 -1
  39. mteb/models/model_implementations/emillykkejensen_models.py +3 -3
  40. mteb/models/model_implementations/en_code_retriever.py +1 -1
  41. mteb/models/model_implementations/euler_models.py +2 -2
  42. mteb/models/model_implementations/fa_models.py +9 -9
  43. mteb/models/model_implementations/facebookai.py +14 -2
  44. mteb/models/model_implementations/geogpt_models.py +1 -1
  45. mteb/models/model_implementations/gme_v_models.py +2 -2
  46. mteb/models/model_implementations/google_models.py +1 -1
  47. mteb/models/model_implementations/granite_vision_embedding_models.py +1 -1
  48. mteb/models/model_implementations/gritlm_models.py +2 -2
  49. mteb/models/model_implementations/gte_models.py +25 -13
  50. mteb/models/model_implementations/hinvec_models.py +1 -1
  51. mteb/models/model_implementations/ibm_granite_models.py +30 -6
  52. mteb/models/model_implementations/inf_models.py +2 -2
  53. mteb/models/model_implementations/jasper_models.py +2 -2
  54. mteb/models/model_implementations/jina_clip.py +1 -1
  55. mteb/models/model_implementations/jina_models.py +11 -5
  56. mteb/models/model_implementations/kblab.py +12 -6
  57. mteb/models/model_implementations/kennethenevoldsen_models.py +2 -2
  58. mteb/models/model_implementations/kfst.py +1 -1
  59. mteb/models/model_implementations/kowshik24_models.py +1 -1
  60. mteb/models/model_implementations/lgai_embedding_models.py +1 -1
  61. mteb/models/model_implementations/linq_models.py +1 -1
  62. mteb/models/model_implementations/listconranker.py +1 -1
  63. mteb/models/model_implementations/llm2clip_models.py +3 -3
  64. mteb/models/model_implementations/llm2vec_models.py +8 -8
  65. mteb/models/model_implementations/mdbr_models.py +14 -2
  66. mteb/models/model_implementations/misc_models.py +68 -68
  67. mteb/models/model_implementations/mme5_models.py +1 -1
  68. mteb/models/model_implementations/moco_models.py +2 -2
  69. mteb/models/model_implementations/mod_models.py +1 -1
  70. mteb/models/model_implementations/model2vec_models.py +13 -13
  71. mteb/models/model_implementations/moka_models.py +1 -1
  72. mteb/models/model_implementations/mxbai_models.py +16 -3
  73. mteb/models/model_implementations/nbailab.py +3 -3
  74. mteb/models/model_implementations/no_instruct_sentence_models.py +1 -1
  75. mteb/models/model_implementations/nomic_models.py +18 -6
  76. mteb/models/model_implementations/nomic_models_vision.py +1 -1
  77. mteb/models/model_implementations/nvidia_llama_nemoretriever_colemb.py +2 -2
  78. mteb/models/model_implementations/nvidia_models.py +3 -3
  79. mteb/models/model_implementations/octen_models.py +3 -3
  80. mteb/models/model_implementations/openclip_models.py +6 -6
  81. mteb/models/model_implementations/opensearch_neural_sparse_models.py +5 -5
  82. mteb/models/model_implementations/ops_moa_models.py +1 -1
  83. mteb/models/model_implementations/ordalietech_solon_embeddings_mini_beta_1_1.py +1 -1
  84. mteb/models/model_implementations/pawan_models.py +1 -1
  85. mteb/models/model_implementations/piccolo_models.py +1 -1
  86. mteb/models/model_implementations/promptriever_models.py +4 -4
  87. mteb/models/model_implementations/pylate_models.py +5 -5
  88. mteb/models/model_implementations/qodo_models.py +2 -2
  89. mteb/models/model_implementations/qtack_models.py +1 -1
  90. mteb/models/model_implementations/qwen3_models.py +3 -3
  91. mteb/models/model_implementations/qzhou_models.py +2 -2
  92. mteb/models/model_implementations/rasgaard_models.py +1 -1
  93. mteb/models/model_implementations/reasonir_model.py +1 -1
  94. mteb/models/model_implementations/repllama_models.py +1 -1
  95. mteb/models/model_implementations/rerankers_custom.py +9 -3
  96. mteb/models/model_implementations/rerankers_monot5_based.py +14 -14
  97. mteb/models/model_implementations/richinfoai_models.py +1 -1
  98. mteb/models/model_implementations/ru_sentence_models.py +20 -20
  99. mteb/models/model_implementations/ruri_models.py +10 -10
  100. mteb/models/model_implementations/salesforce_models.py +3 -3
  101. mteb/models/model_implementations/samilpwc_models.py +1 -1
  102. mteb/models/model_implementations/sarashina_embedding_models.py +2 -2
  103. mteb/models/model_implementations/searchmap_models.py +1 -1
  104. mteb/models/model_implementations/sentence_transformers_models.py +58 -22
  105. mteb/models/model_implementations/shuu_model.py +1 -1
  106. mteb/models/model_implementations/siglip_models.py +10 -10
  107. mteb/models/model_implementations/slm_models.py +416 -0
  108. mteb/models/model_implementations/spartan8806_atles_champion.py +1 -1
  109. mteb/models/model_implementations/stella_models.py +17 -4
  110. mteb/models/model_implementations/tarka_models.py +2 -2
  111. mteb/models/model_implementations/text2vec_models.py +9 -3
  112. mteb/models/model_implementations/ua_sentence_models.py +1 -1
  113. mteb/models/model_implementations/uae_models.py +7 -1
  114. mteb/models/model_implementations/vdr_models.py +1 -1
  115. mteb/models/model_implementations/vi_vn_models.py +6 -6
  116. mteb/models/model_implementations/vlm2vec_models.py +2 -2
  117. mteb/models/model_implementations/youtu_models.py +1 -1
  118. mteb/models/model_implementations/yuan_models.py +1 -1
  119. mteb/models/model_implementations/yuan_models_en.py +1 -1
  120. mteb/models/model_meta.py +46 -17
  121. mteb/results/benchmark_results.py +2 -2
  122. mteb/tasks/classification/kur/kurdish_sentiment_classification.py +2 -2
  123. mteb/tasks/clustering/eng/hume_wiki_cities_clustering.py +1 -1
  124. mteb/tasks/clustering/eng/wiki_cities_clustering.py +1 -1
  125. mteb/tasks/clustering/zho/cmteb_clustering.py +2 -2
  126. mteb/tasks/reranking/multilingual/wikipedia_reranking_multilingual.py +1 -1
  127. mteb/tasks/retrieval/eng/cub200_i2i_retrieval.py +1 -1
  128. mteb/tasks/retrieval/vie/__init__.py +14 -6
  129. mteb/tasks/retrieval/vie/climate_fevervn_retrieval.py +39 -0
  130. mteb/tasks/retrieval/vie/db_pedia_vn_retrieval.py +39 -0
  131. mteb/tasks/retrieval/vie/fevervn_retrieval.py +39 -0
  132. mteb/tasks/retrieval/vie/hotpot_qavn_retrieval.py +39 -0
  133. mteb/tasks/retrieval/vie/msmarcovn_retrieval.py +48 -0
  134. mteb/tasks/retrieval/vie/nqvn_retrieval.py +39 -0
  135. mteb/tasks/retrieval/vie/tvpl_retrieval.py +42 -0
  136. mteb/tasks/retrieval/vie/zac_legal_text_retrieval.py +15 -1
  137. {mteb-2.6.4.dist-info → mteb-2.6.6.dist-info}/METADATA +3 -3
  138. {mteb-2.6.4.dist-info → mteb-2.6.6.dist-info}/RECORD +142 -133
  139. {mteb-2.6.4.dist-info → mteb-2.6.6.dist-info}/WHEEL +0 -0
  140. {mteb-2.6.4.dist-info → mteb-2.6.6.dist-info}/entry_points.txt +0 -0
  141. {mteb-2.6.4.dist-info → mteb-2.6.6.dist-info}/licenses/LICENSE +0 -0
  142. {mteb-2.6.4.dist-info → mteb-2.6.6.dist-info}/top_level.txt +0 -0
mteb/models/model_implementations/colsmol_models.py CHANGED
@@ -67,7 +67,7 @@ colsmol_256m = ModelMeta(
67
67
  open_weights=True,
68
68
  public_training_code="https://github.com/illuin-tech/colpali",
69
69
  public_training_data="https://huggingface.co/datasets/vidore/colpali_train_set",
70
- framework=["ColPali"],
70
+ framework=["ColPali", "safetensors"],
71
71
  reference="https://huggingface.co/vidore/colSmol-256M",
72
72
  similarity_fn_name="MaxSim",
73
73
  use_instructions=True,
@@ -94,7 +94,7 @@ colsmol_500m = ModelMeta(
94
94
  open_weights=True,
95
95
  public_training_code="https://github.com/illuin-tech/colpali",
96
96
  public_training_data="https://huggingface.co/datasets/vidore/colpali_train_set",
97
- framework=["ColPali"],
97
+ framework=["ColPali", "safetensors"],
98
98
  reference="https://huggingface.co/vidore/colSmol-500M",
99
99
  similarity_fn_name="MaxSim",
100
100
  use_instructions=True,
mteb/models/model_implementations/conan_models.py CHANGED
@@ -209,7 +209,7 @@ Conan_embedding_v2 = ModelMeta(
209
209
  license="apache-2.0",
210
210
  reference="https://huggingface.co/TencentBAC/Conan-embedding-v2",
211
211
  similarity_fn_name="cosine",
212
- framework=["API"],
212
+ framework=["API", "Sentence Transformers", "Transformers"],
213
213
  use_instructions=True,
214
214
  training_datasets=E5_MISTRAL_TRAINING_DATA | bge_full_data | conan_zh_datasets,
215
215
  public_training_code=None,
mteb/models/model_implementations/dino_models.py CHANGED
@@ -119,7 +119,7 @@ dinov2_small = ModelMeta(
119
119
  open_weights=True,
120
120
  public_training_code="https://github.com/facebookresearch/dinov2",
121
121
  public_training_data=None,
122
- framework=["PyTorch"],
122
+ framework=["PyTorch", "Transformers", "safetensors"],
123
123
  reference="https://huggingface.co/facebook/dinov2-small",
124
124
  similarity_fn_name=ScoringFunction.COSINE,
125
125
  use_instructions=False,
@@ -150,7 +150,7 @@ dinov2_base = ModelMeta(
150
150
  open_weights=True,
151
151
  public_training_code="https://github.com/facebookresearch/dinov2",
152
152
  public_training_data=None,
153
- framework=["PyTorch"],
153
+ framework=["PyTorch", "Transformers", "safetensors"],
154
154
  reference="https://huggingface.co/facebook/dinov2-base",
155
155
  similarity_fn_name=ScoringFunction.COSINE,
156
156
  use_instructions=False,
@@ -181,7 +181,7 @@ dinov2_large = ModelMeta(
181
181
  open_weights=True,
182
182
  public_training_code="https://github.com/facebookresearch/dinov2",
183
183
  public_training_data=None,
184
- framework=["PyTorch"],
184
+ framework=["PyTorch", "Transformers", "safetensors"],
185
185
  reference="https://huggingface.co/facebook/dinov2-large",
186
186
  similarity_fn_name=ScoringFunction.COSINE,
187
187
  use_instructions=False,
@@ -212,7 +212,7 @@ dinov2_giant = ModelMeta(
212
212
  open_weights=True,
213
213
  public_training_code="https://github.com/facebookresearch/dinov2",
214
214
  public_training_data=None,
215
- framework=["PyTorch"],
215
+ framework=["PyTorch", "Transformers", "safetensors"],
216
216
  reference="https://huggingface.co/facebook/dinov2-giant",
217
217
  similarity_fn_name=ScoringFunction.COSINE,
218
218
  use_instructions=False,
@@ -247,7 +247,7 @@ webssl_dino300m_full2b = ModelMeta(
247
247
  open_weights=True,
248
248
  public_training_code="",
249
249
  public_training_data=None,
250
- framework=["PyTorch"],
250
+ framework=["PyTorch", "Transformers", "safetensors"],
251
251
  reference="https://huggingface.co/facebook/webssl-dino300m-full2b-224",
252
252
  similarity_fn_name=None,
253
253
  use_instructions=False,
@@ -278,7 +278,7 @@ webssl_dino1b_full2b = ModelMeta(
278
278
  open_weights=True,
279
279
  public_training_code="",
280
280
  public_training_data=None,
281
- framework=["PyTorch"],
281
+ framework=["PyTorch", "Transformers", "safetensors"],
282
282
  reference="https://huggingface.co/facebook/webssl-dino1b-full2b-224",
283
283
  similarity_fn_name=None,
284
284
  use_instructions=False,
@@ -309,7 +309,7 @@ webssl_dino2b_full2b = ModelMeta(
309
309
  open_weights=True,
310
310
  public_training_code="",
311
311
  public_training_data=None,
312
- framework=["PyTorch"],
312
+ framework=["PyTorch", "Transformers", "safetensors"],
313
313
  reference="https://huggingface.co/facebook/webssl-dino2b-full2b-224",
314
314
  similarity_fn_name=None,
315
315
  use_instructions=False,
@@ -340,7 +340,7 @@ webssl_dino3b_full2b = ModelMeta(
340
340
  open_weights=True,
341
341
  public_training_code="",
342
342
  public_training_data=None,
343
- framework=["PyTorch"],
343
+ framework=["PyTorch", "Transformers", "safetensors"],
344
344
  reference="https://huggingface.co/facebook/webssl-dino3b-full2b-224",
345
345
  similarity_fn_name=None,
346
346
  use_instructions=False,
@@ -371,7 +371,7 @@ webssl_dino5b_full2b = ModelMeta(
371
371
  open_weights=True,
372
372
  public_training_code="",
373
373
  public_training_data=None,
374
- framework=["PyTorch"],
374
+ framework=["PyTorch", "Transformers", "safetensors"],
375
375
  reference="https://huggingface.co/facebook/webssl-dino5b-full2b-224",
376
376
  similarity_fn_name=None,
377
377
  use_instructions=False,
@@ -402,7 +402,7 @@ webssl_dino7b_full8b_224 = ModelMeta(
402
402
  open_weights=True,
403
403
  public_training_code="",
404
404
  public_training_data=None,
405
- framework=["PyTorch"],
405
+ framework=["PyTorch", "Transformers", "safetensors"],
406
406
  reference="https://huggingface.co/facebook/webssl-dino7b-full8b-224",
407
407
  similarity_fn_name=None,
408
408
  use_instructions=False,
@@ -433,7 +433,7 @@ webssl_dino7b_full8b_378 = ModelMeta(
433
433
  open_weights=True,
434
434
  public_training_code="",
435
435
  public_training_data=None,
436
- framework=["PyTorch"],
436
+ framework=["PyTorch", "Transformers", "safetensors"],
437
437
  reference="https://huggingface.co/facebook/webssl-dino7b-full8b-378",
438
438
  similarity_fn_name=None,
439
439
  use_instructions=False,
@@ -464,7 +464,7 @@ webssl_dino7b_full8b_518 = ModelMeta(
464
464
  open_weights=True,
465
465
  public_training_code="",
466
466
  public_training_data=None,
467
- framework=["PyTorch"],
467
+ framework=["PyTorch", "Transformers", "safetensors"],
468
468
  reference="https://huggingface.co/facebook/webssl-dino7b-full8b-518",
469
469
  similarity_fn_name=None,
470
470
  use_instructions=False,
@@ -496,7 +496,7 @@ webssl_dino2b_light2b = ModelMeta(
496
496
  open_weights=True,
497
497
  public_training_code="",
498
498
  public_training_data=None,
499
- framework=["PyTorch"],
499
+ framework=["PyTorch", "Transformers", "safetensors"],
500
500
  reference="https://huggingface.co/facebook/webssl-dino2b-light2b-224",
501
501
  similarity_fn_name=None,
502
502
  use_instructions=False,
@@ -527,7 +527,7 @@ webssl_dino2b_heavy2b = ModelMeta(
527
527
  open_weights=True,
528
528
  public_training_code="",
529
529
  public_training_data=None,
530
- framework=["PyTorch"],
530
+ framework=["PyTorch", "Transformers", "safetensors"],
531
531
  reference="https://huggingface.co/facebook/webssl-dino2b-heavy2b-224",
532
532
  similarity_fn_name=None,
533
533
  use_instructions=False,
@@ -558,7 +558,7 @@ webssl_dino3b_light2b = ModelMeta(
558
558
  open_weights=True,
559
559
  public_training_code="",
560
560
  public_training_data=None,
561
- framework=["PyTorch"],
561
+ framework=["PyTorch", "Transformers", "safetensors"],
562
562
  reference="https://huggingface.co/facebook/webssl-dino3b-light2b-224",
563
563
  similarity_fn_name=None,
564
564
  use_instructions=False,
@@ -589,7 +589,7 @@ webssl_dino3b_heavy2b = ModelMeta(
589
589
  open_weights=True,
590
590
  public_training_code="",
591
591
  public_training_data=None,
592
- framework=["PyTorch"],
592
+ framework=["PyTorch", "Transformers", "safetensors"],
593
593
  reference="https://huggingface.co/facebook/webssl-dino3b-heavy2b-224",
594
594
  similarity_fn_name=None,
595
595
  use_instructions=False,
@@ -620,7 +620,7 @@ webssl_mae300m_full2b = ModelMeta(
620
620
  open_weights=True,
621
621
  public_training_code="",
622
622
  public_training_data=None,
623
- framework=["PyTorch"],
623
+ framework=["PyTorch", "Transformers", "safetensors"],
624
624
  reference="https://huggingface.co/facebook/webssl-mae300m-full2b-224",
625
625
  similarity_fn_name=None,
626
626
  use_instructions=False,
@@ -651,7 +651,7 @@ webssl_mae700m_full2b = ModelMeta(
651
651
  open_weights=True,
652
652
  public_training_code="",
653
653
  public_training_data=None,
654
- framework=["PyTorch"],
654
+ framework=["PyTorch", "Transformers", "safetensors"],
655
655
  reference="https://huggingface.co/facebook/webssl-mae700m-full2b-224",
656
656
  similarity_fn_name=None,
657
657
  use_instructions=False,
@@ -682,7 +682,7 @@ webssl_mae1b_full2b = ModelMeta(
682
682
  open_weights=True,
683
683
  public_training_code="",
684
684
  public_training_data=None,
685
- framework=["PyTorch"],
685
+ framework=["PyTorch", "Transformers", "safetensors"],
686
686
  reference="https://huggingface.co/facebook/webssl-mae1b-full2b-224",
687
687
  similarity_fn_name=None,
688
688
  use_instructions=False,
mteb/models/model_implementations/e5_instruct.py CHANGED
@@ -45,7 +45,14 @@ e5_instruct = ModelMeta(
45
45
  open_weights=True,
46
46
  revision="baa7be480a7de1539afce709c8f13f833a510e0a",
47
47
  release_date=E5_PAPER_RELEASE_DATE,
48
- framework=["GritLM", "PyTorch", "Sentence Transformers"],
48
+ framework=[
49
+ "GritLM",
50
+ "PyTorch",
51
+ "Sentence Transformers",
52
+ "ONNX",
53
+ "safetensors",
54
+ "Transformers",
55
+ ],
49
56
  similarity_fn_name=ScoringFunction.COSINE,
50
57
  use_instructions=True,
51
58
  reference="https://huggingface.co/intfloat/multilingual-e5-large-instruct",
@@ -84,7 +91,13 @@ e5_mistral = ModelMeta(
84
91
  open_weights=True,
85
92
  revision="07163b72af1488142a360786df853f237b1a3ca1",
86
93
  release_date=E5_PAPER_RELEASE_DATE,
87
- framework=["GritLM", "PyTorch", "Sentence Transformers"],
94
+ framework=[
95
+ "GritLM",
96
+ "PyTorch",
97
+ "Sentence Transformers",
98
+ "safetensors",
99
+ "Transformers",
100
+ ],
88
101
  similarity_fn_name=ScoringFunction.COSINE,
89
102
  use_instructions=True,
90
103
  reference="https://huggingface.co/intfloat/e5-mistral-7b-instruct",
@@ -139,7 +152,13 @@ zeta_alpha_ai__zeta_alpha_e5_mistral = ModelMeta(
139
152
  open_weights=True,
140
153
  public_training_data=None,
141
154
  public_training_code=None,
142
- framework=["PyTorch", "Sentence Transformers", "GritLM"],
155
+ framework=[
156
+ "PyTorch",
157
+ "Sentence Transformers",
158
+ "GritLM",
159
+ "safetensors",
160
+ "Transformers",
161
+ ],
143
162
  reference="https://huggingface.co/zeta-alpha-ai/Zeta-Alpha-E5-Mistral",
144
163
  similarity_fn_name=ScoringFunction.COSINE,
145
164
  use_instructions=True,
@@ -216,7 +235,7 @@ BeastyZ__e5_R_mistral_7b = ModelMeta(
216
235
  open_weights=True,
217
236
  public_training_code="https://github.com/LeeSureman/E5-Retrieval-Reproduction",
218
237
  public_training_data="https://huggingface.co/datasets/BeastyZ/E5-R",
219
- framework=["PyTorch"],
238
+ framework=["PyTorch", "Transformers", "safetensors"],
220
239
  reference="https://huggingface.co/BeastyZ/e5-R-mistral-7b",
221
240
  similarity_fn_name="cosine",
222
241
  use_instructions=True,
mteb/models/model_implementations/e5_models.py CHANGED
@@ -82,7 +82,7 @@ e5_mult_small = ModelMeta(
82
82
  max_tokens=512,
83
83
  reference="https://huggingface.co/intfloat/multilingual-e5-small",
84
84
  similarity_fn_name=ScoringFunction.COSINE,
85
- framework=["Sentence Transformers", "PyTorch"],
85
+ framework=["Sentence Transformers", "PyTorch", "ONNX", "safetensors"],
86
86
  use_instructions=True,
87
87
  public_training_code=None, # couldn't find
88
88
  public_training_data=None,
@@ -109,7 +109,7 @@ e5_mult_base = ModelMeta(
109
109
  max_tokens=514,
110
110
  reference="https://huggingface.co/intfloat/multilingual-e5-base",
111
111
  similarity_fn_name=ScoringFunction.COSINE,
112
- framework=["Sentence Transformers", "PyTorch"],
112
+ framework=["Sentence Transformers", "PyTorch", "ONNX", "safetensors"],
113
113
  use_instructions=True,
114
114
  public_training_code=None,
115
115
  public_training_data=None,
@@ -136,7 +136,7 @@ e5_mult_large = ModelMeta(
136
136
  max_tokens=514,
137
137
  reference="https://huggingface.co/intfloat/multilingual-e5-large",
138
138
  similarity_fn_name=ScoringFunction.COSINE,
139
- framework=["Sentence Transformers", "PyTorch"],
139
+ framework=["Sentence Transformers", "PyTorch", "ONNX", "safetensors"],
140
140
  use_instructions=True,
141
141
  public_training_code=None,
142
142
  public_training_data=None,
@@ -163,7 +163,7 @@ e5_eng_small_v2 = ModelMeta(
163
163
  max_tokens=512,
164
164
  reference="https://huggingface.co/intfloat/e5-small-v2",
165
165
  similarity_fn_name=ScoringFunction.COSINE,
166
- framework=["Sentence Transformers", "PyTorch"],
166
+ framework=["Sentence Transformers", "PyTorch", "ONNX", "safetensors"],
167
167
  use_instructions=True,
168
168
  public_training_code=None,
169
169
  public_training_data=None,
@@ -190,7 +190,7 @@ e5_eng_small = ModelMeta(
190
190
  max_tokens=512,
191
191
  reference="https://huggingface.co/intfloat/e5-small",
192
192
  similarity_fn_name=ScoringFunction.COSINE,
193
- framework=["Sentence Transformers", "PyTorch"],
193
+ framework=["Sentence Transformers", "PyTorch", "ONNX", "safetensors"],
194
194
  use_instructions=True,
195
195
  public_training_code=None,
196
196
  public_training_data=None,
@@ -217,7 +217,7 @@ e5_eng_base_v2 = ModelMeta(
217
217
  max_tokens=512,
218
218
  reference="https://huggingface.co/intfloat/e5-base-v2",
219
219
  similarity_fn_name=ScoringFunction.COSINE,
220
- framework=["Sentence Transformers", "PyTorch"],
220
+ framework=["Sentence Transformers", "PyTorch", "ONNX", "safetensors"],
221
221
  use_instructions=True,
222
222
  superseded_by=None,
223
223
  adapted_from="intfloat/e5-base",
@@ -245,7 +245,7 @@ e5_eng_large_v2 = ModelMeta(
245
245
  max_tokens=514,
246
246
  reference="https://huggingface.co/intfloat/e5-large-v2",
247
247
  similarity_fn_name=ScoringFunction.COSINE,
248
- framework=["Sentence Transformers", "PyTorch"],
248
+ framework=["Sentence Transformers", "PyTorch", "ONNX", "safetensors"],
249
249
  use_instructions=True,
250
250
  superseded_by=None,
251
251
  adapted_from="intfloat/e5-large",
@@ -273,7 +273,7 @@ e5_large = ModelMeta(
273
273
  max_tokens=512,
274
274
  reference="https://huggingface.co/intfloat/e5-large",
275
275
  similarity_fn_name=ScoringFunction.COSINE,
276
- framework=["Sentence Transformers", "PyTorch"],
276
+ framework=["Sentence Transformers", "PyTorch", "safetensors"],
277
277
  use_instructions=True,
278
278
  superseded_by="intfloat/e5-large-v2",
279
279
  adapted_from="google-bert/bert-large-uncased-whole-word-masking",
@@ -301,7 +301,7 @@ e5_base = ModelMeta(
301
301
  max_tokens=512,
302
302
  reference="https://huggingface.co/intfloat/e5-base",
303
303
  similarity_fn_name=ScoringFunction.COSINE,
304
- framework=["Sentence Transformers", "PyTorch"],
304
+ framework=["Sentence Transformers", "PyTorch", "safetensors"],
305
305
  use_instructions=True,
306
306
  superseded_by="intfloat/e5-base-v2",
307
307
  adapted_from="google-bert/bert-base-uncased",
mteb/models/model_implementations/e5_v.py CHANGED
@@ -173,7 +173,7 @@ e5_v = ModelMeta(
173
173
  open_weights=True,
174
174
  public_training_code="https://github.com/kongds/E5-V",
175
175
  public_training_data="https://huggingface.co/datasets/princeton-nlp/datasets-for-simcse",
176
- framework=["PyTorch"],
176
+ framework=["PyTorch", "Transformers", "safetensors"],
177
177
  reference="https://huggingface.co/royokong/e5-v",
178
178
  similarity_fn_name=ScoringFunction.COSINE,
179
179
  use_instructions=True,
mteb/models/model_implementations/eagerworks_models.py CHANGED
@@ -152,7 +152,7 @@ Eager_Embed_V1 = ModelMeta(
152
152
  embed_dim=2560,
153
153
  license="apache-2.0",
154
154
  open_weights=True,
155
- framework=["Tevatron"],
155
+ framework=["Tevatron", "safetensors"],
156
156
  reference="https://huggingface.co/eagerworks/eager-embed-v1",
157
157
  similarity_fn_name=ScoringFunction.COSINE,
158
158
  use_instructions=True,
mteb/models/model_implementations/emillykkejensen_models.py CHANGED
@@ -14,7 +14,7 @@ embedding_gemma_300m_scandi = ModelMeta(
14
14
  max_tokens=2048,
15
15
  license="apache-2.0",
16
16
  reference="https://huggingface.co/emillykkejensen/EmbeddingGemma-Scandi-300m",
17
- framework=["Sentence Transformers", "PyTorch"],
17
+ framework=["Sentence Transformers", "PyTorch", "safetensors"],
18
18
  use_instructions=True,
19
19
  public_training_code=None,
20
20
  public_training_data="https://huggingface.co/datasets/DDSC/nordic-embedding-training-data",
@@ -48,7 +48,7 @@ qwen_scandi = ModelMeta(
48
48
  max_tokens=32768,
49
49
  license="apache-2.0",
50
50
  reference="https://huggingface.co/emillykkejensen/Qwen3-Embedding-Scandi-0.6B",
51
- framework=["Sentence Transformers", "PyTorch"],
51
+ framework=["Sentence Transformers", "PyTorch", "safetensors"],
52
52
  use_instructions=True,
53
53
  public_training_code=None,
54
54
  public_training_data="https://huggingface.co/datasets/DDSC/nordic-embedding-training-data",
@@ -72,7 +72,7 @@ mmbert_scandi = ModelMeta(
72
72
  max_tokens=8192,
73
73
  license="apache-2.0",
74
74
  reference="https://huggingface.co/emillykkejensen/Qwen3-Embedding-Scandi-0.6B",
75
- framework=["Sentence Transformers", "PyTorch"],
75
+ framework=["Sentence Transformers", "PyTorch", "safetensors"],
76
76
  use_instructions=True,
77
77
  public_training_code=None,
78
78
  public_training_data="https://huggingface.co/datasets/DDSC/nordic-embedding-training-data",
mteb/models/model_implementations/en_code_retriever.py CHANGED
@@ -24,7 +24,7 @@ english_code_retriever = ModelMeta(
24
24
  max_tokens=8192,
25
25
  reference="https://huggingface.co/fyaronskiy/english_code_retriever",
26
26
  similarity_fn_name="cosine",
27
- framework=["Sentence Transformers", "PyTorch"],
27
+ framework=["Sentence Transformers", "PyTorch", "safetensors"],
28
28
  use_instructions=True,
29
29
  public_training_code=None,
30
30
  public_training_data="https://huggingface.co/datasets/code-search-net/code_search_net",
mteb/models/model_implementations/euler_models.py CHANGED
@@ -16,7 +16,7 @@ Euler_Legal_Embedding_V1 = ModelMeta(
16
16
  open_weights=True,
17
17
  public_training_code=None,
18
18
  public_training_data=None,
19
- framework=["PyTorch", "Sentence Transformers"],
19
+ framework=["PyTorch", "Sentence Transformers", "safetensors", "Transformers"],
20
20
  reference="https://huggingface.co/Mira190/Euler-Legal-Embedding-V1",
21
21
  similarity_fn_name="cosine",
22
22
  use_instructions=False,
@@ -24,7 +24,7 @@ Euler_Legal_Embedding_V1 = ModelMeta(
24
24
  adapted_from="Qwen/Qwen3-Embedding-8B",
25
25
  superseded_by=None,
26
26
  citation="""@misc{euler2025legal,
27
- title={Euler-Legal-Embedding: Advanced Legal Representation Learning},
27
+ title={Euler-Legal-Embedding: Advanced Legal Representation Learning},
28
28
  author={LawRank Team},
29
29
  year={2025},
30
30
  publisher={Hugging Face}
mteb/models/model_implementations/fa_models.py CHANGED
@@ -18,7 +18,7 @@ parsbert = ModelMeta(
18
18
  max_tokens=512,
19
19
  reference="https://huggingface.co/HooshvareLab/bert-base-parsbert-uncased",
20
20
  similarity_fn_name=ScoringFunction.COSINE,
21
- framework=["Sentence Transformers", "PyTorch"],
21
+ framework=["Sentence Transformers", "PyTorch", "Transformers"],
22
22
  use_instructions=False,
23
23
  public_training_code=None,
24
24
  public_training_data=None,
@@ -54,7 +54,7 @@ bert_zwnj = ModelMeta(
54
54
  max_tokens=512,
55
55
  reference="https://huggingface.co/m3hrdadfi/bert-zwnj-wnli-mean-tokens",
56
56
  similarity_fn_name=ScoringFunction.COSINE,
57
- framework=["Sentence Transformers", "PyTorch"],
57
+ framework=["Sentence Transformers", "PyTorch", "Transformers"],
58
58
  use_instructions=False,
59
59
  public_training_code=None,
60
60
  public_training_data=None,
@@ -80,7 +80,7 @@ roberta_zwnj = ModelMeta(
80
80
  max_tokens=514,
81
81
  reference="https://huggingface.co/m3hrdadfi/roberta-zwnj-wnli-mean-tokens",
82
82
  similarity_fn_name=ScoringFunction.COSINE,
83
- framework=["Sentence Transformers", "PyTorch"],
83
+ framework=["Sentence Transformers", "PyTorch", "Transformers"],
84
84
  use_instructions=False,
85
85
  public_training_code=None,
86
86
  public_training_data=None,
@@ -105,7 +105,7 @@ sentence_transformer_parsbert = ModelMeta(
105
105
  max_tokens=512,
106
106
  reference="https://huggingface.co/myrkur/sentence-transformer-parsbert-fa",
107
107
  similarity_fn_name=ScoringFunction.COSINE,
108
- framework=["Sentence Transformers", "PyTorch"],
108
+ framework=["Sentence Transformers", "PyTorch", "safetensors"],
109
109
  use_instructions=False,
110
110
  public_training_code=None,
111
111
  public_training_data=None,
@@ -129,7 +129,7 @@ tooka_bert_base = ModelMeta(
129
129
  max_tokens=512,
130
130
  reference="https://huggingface.co/PartAI/TookaBERT-Base",
131
131
  similarity_fn_name=ScoringFunction.COSINE,
132
- framework=["Sentence Transformers", "PyTorch"],
132
+ framework=["Sentence Transformers", "PyTorch", "Transformers", "safetensors"],
133
133
  use_instructions=False,
134
134
  public_training_code=None,
135
135
  public_training_data=None,
@@ -156,7 +156,7 @@ tooka_sbert = ModelMeta(
156
156
  max_tokens=512,
157
157
  reference="https://huggingface.co/PartAI/Tooka-SBERT",
158
158
  similarity_fn_name=ScoringFunction.COSINE,
159
- framework=["Sentence Transformers", "PyTorch"],
159
+ framework=["Sentence Transformers", "PyTorch", "safetensors"],
160
160
  use_instructions=False,
161
161
  public_training_code=None,
162
162
  public_training_data=None,
@@ -187,7 +187,7 @@ fa_bert = ModelMeta(
187
187
  max_tokens=512,
188
188
  reference="https://huggingface.co/sbunlp/fabert",
189
189
  similarity_fn_name=ScoringFunction.COSINE,
190
- framework=["Sentence Transformers", "PyTorch"],
190
+ framework=["Sentence Transformers", "PyTorch", "Transformers", "safetensors"],
191
191
  use_instructions=False,
192
192
  public_training_code=None,
193
193
  public_training_data=None,
@@ -235,7 +235,7 @@ tooka_sbert_v2_small = ModelMeta(
235
235
  max_tokens=512,
236
236
  reference="https://huggingface.co/PartAI/Tooka-SBERT-V2-Small",
237
237
  similarity_fn_name="cosine",
238
- framework=["Sentence Transformers", "PyTorch"],
238
+ framework=["Sentence Transformers", "PyTorch", "safetensors"],
239
239
  use_instructions=False,
240
240
  public_training_code=None,
241
241
  public_training_data=None,
@@ -266,7 +266,7 @@ tooka_sbert_v2_large = ModelMeta(
266
266
  max_tokens=512,
267
267
  reference="https://huggingface.co/PartAI/Tooka-SBERT-V2-Large",
268
268
  similarity_fn_name="cosine",
269
- framework=["Sentence Transformers", "PyTorch"],
269
+ framework=["Sentence Transformers", "PyTorch", "safetensors"],
270
270
  use_instructions=False,
271
271
  public_training_code=None,
272
272
  public_training_data=None,
mteb/models/model_implementations/facebookai.py CHANGED
@@ -119,7 +119,13 @@ xlmr_base = ModelMeta(
119
119
  max_tokens=512,
120
120
  reference="https://huggingface.co/FacebookAI/xlm-roberta-base",
121
121
  similarity_fn_name=ScoringFunction.COSINE,
122
- framework=["Sentence Transformers", "PyTorch"],
122
+ framework=[
123
+ "Sentence Transformers",
124
+ "PyTorch",
125
+ "Transformers",
126
+ "ONNX",
127
+ "safetensors",
128
+ ],
123
129
  use_instructions=False,
124
130
  public_training_code=None,
125
131
  public_training_data=None,
@@ -163,7 +169,13 @@ xlmr_large = ModelMeta(
163
169
  max_tokens=512,
164
170
  reference="https://huggingface.co/FacebookAI/xlm-roberta-large",
165
171
  similarity_fn_name=ScoringFunction.COSINE,
166
- framework=["Sentence Transformers", "PyTorch"],
172
+ framework=[
173
+ "Sentence Transformers",
174
+ "PyTorch",
175
+ "Transformers",
176
+ "ONNX",
177
+ "safetensors",
178
+ ],
167
179
  use_instructions=False,
168
180
  public_training_code=None,
169
181
  public_training_data=None,
mteb/models/model_implementations/geogpt_models.py CHANGED
@@ -26,7 +26,7 @@ geoembedding = ModelMeta(
26
26
  max_tokens=32768,
27
27
  reference="https://huggingface.co/GeoGPT-Research-Project/GeoEmbedding",
28
28
  similarity_fn_name="cosine",
29
- framework=["Sentence Transformers", "PyTorch"],
29
+ framework=["Sentence Transformers", "PyTorch", "safetensors", "Transformers"],
30
30
  use_instructions=True,
31
31
  public_training_code=None,
32
32
  public_training_data=None,
mteb/models/model_implementations/gme_v_models.py CHANGED
@@ -360,7 +360,7 @@ gme_qwen2vl_2b = ModelMeta(
360
360
  max_tokens=32768,
361
361
  reference="https://huggingface.co/Alibaba-NLP/gme-Qwen2-VL-2B-Instruct",
362
362
  similarity_fn_name=ScoringFunction.COSINE,
363
- framework=["PyTorch"],
363
+ framework=["PyTorch", "Sentence Transformers", "safetensors", "Transformers"],
364
364
  use_instructions=True,
365
365
  public_training_code=None,
366
366
  public_training_data=None,
@@ -384,7 +384,7 @@ gme_qwen2vl_7b = ModelMeta(
384
384
  max_tokens=32768,
385
385
  reference="https://huggingface.co/Alibaba-NLP/gme-Qwen2-VL-7B-Instruct",
386
386
  similarity_fn_name=ScoringFunction.COSINE,
387
- framework=["PyTorch"],
387
+ framework=["PyTorch", "Sentence Transformers", "safetensors", "Transformers"],
388
388
  use_instructions=True,
389
389
  public_training_code=None,
390
390
  public_training_data=None,
mteb/models/model_implementations/google_models.py CHANGED
@@ -270,7 +270,7 @@ embedding_gemma_300m = ModelMeta(
270
270
  max_tokens=2048,
271
271
  license="gemma",
272
272
  reference="https://ai.google.dev/gemma/docs/embeddinggemma/model_card",
273
- framework=["Sentence Transformers", "PyTorch"],
273
+ framework=["Sentence Transformers", "PyTorch", "safetensors"],
274
274
  use_instructions=True,
275
275
  public_training_code=None,
276
276
  public_training_data=None,
mteb/models/model_implementations/granite_vision_embedding_models.py CHANGED
@@ -179,7 +179,7 @@ granite_vision_embedding = ModelMeta(
179
179
  open_weights=True,
180
180
  public_training_code=None,
181
181
  public_training_data=None,
182
- framework=["PyTorch"],
182
+ framework=["PyTorch", "Transformers", "safetensors"],
183
183
  reference="https://huggingface.co/ibm-granite/granite-vision-3.3-2b-embedding",
184
184
  similarity_fn_name="MaxSim",
185
185
  use_instructions=True,
mteb/models/model_implementations/gritlm_models.py CHANGED
@@ -50,7 +50,7 @@ gritlm7b = ModelMeta(
50
50
  max_tokens=32768,
51
51
  reference="https://huggingface.co/GritLM/GritLM-7B",
52
52
  similarity_fn_name=ScoringFunction.COSINE,
53
- framework=["GritLM", "PyTorch"],
53
+ framework=["GritLM", "PyTorch", "Transformers", "safetensors"],
54
54
  use_instructions=True,
55
55
  training_datasets=GRIT_LM_TRAINING_DATA,
56
56
  # section 3.1 "We finetune our final models from Mistral 7B [68] and Mixtral 8x7B [69] using adaptations of E5 [160] and the Tülu 2 data
@@ -79,7 +79,7 @@ gritlm8x7b = ModelMeta(
79
79
  max_tokens=32768,
80
80
  reference="https://huggingface.co/GritLM/GritLM-8x7B",
81
81
  similarity_fn_name=ScoringFunction.COSINE,
82
- framework=["GritLM", "PyTorch"],
82
+ framework=["GritLM", "PyTorch", "Transformers", "safetensors"],
83
83
  use_instructions=True,
84
84
  training_datasets=GRIT_LM_TRAINING_DATA,
85
85
  citation=GRITLM_CITATION,